Incorporating acoustic-phonetic knowledge in hybrid TDNN/HMM frameworks.

Christian Dugast, Laurence Devillers

Browse the full ICASSP paper archive.